asdeasoft.net

Dear STKO,

I am currently conducting a seismic analysis for a cut and cover tunnel that will be exposed to earthquake excitation. In order to simulate the sand material, I have employed the PDMY soil model and calibrated it using the recommended values. However, I have encountered an error during the seismic analysis, specifically, the presence of complex roots.

To resolve this issue, I have attempted various methods and have determined that the error is caused by an improper definition of the c value. Upon closer inspection, I found that when I define C as ranging from 0.3 kPa to 5 kPa, the error with complex roots arises. Conversely, when I define C as 9 kPa, the error is resolved. However, considering that sand is a non-cohesive material, the value of 9 kPa seems inappropriate.

Here is the numerical model:
I would greatly appreciate it if you could review my model and assist me in identifying any potential issues.

Best regards,

Jiaxing Zhou

It happens when internally the material cannot solve the equations for the plasticity. After double-checking your material input, You can try to use smaller time increments, or try the 02 or 03 variants of that material

Thanks for your reply.

I will try to do this.

Best regards,

Jiaxing

I have try to change the material model into PDMY 02, however, this error is not ellimated still. Can you give me some advice more?

Try Penalty handler (1.0e12 1.0e12) because when you have multi point constraints and partitioned mesh, sometimes the transformation may not work as expected.

But probably the most important thing is to use the InitialStateAnalysis on/off for the gravity analyses (since your cohesion is so small, you first need to properly set the gravity stress...)

Try this:

Thanks for your reply, the posted model can be runned stability. However, the material stage remains elastic in the posted numerical model, i.e., the update stage should be assigned to material 3 but not 2. When I do this, the error raised again.

I try to simulate the seismic response using the Manzari Dafalias Material soil model. Could you give me some advices, and is there some example can be adopted as reference?

Best,

Jiaxing

Yes you're right.
However I think there is no way to overcome that issue with that small cohesion (you have now c = 0.0008). After the static elastic gravity you have a portion (over the hole) with positive stresses (much higher than your cohesion) and probably the material model is not able to give a solution for that stress state

Dear STKO,

Thanks for your reply.
It seems that a higher cohesion should be assigned for the zone, and I will try to do this. However, Is it applicable in modeling the sand, i.e., a higher cohesion is defined for the sand? Here, the C = 0.0008 represents that the cohesion is 0.8 kPa. Any suggestion is appreciated.

Best regards,

Jiaxing

Dear STKO,
I want to simulate the soil-tunnel system using PM4Sand model, and I develop two numerical models, i.e., a tunnel soil interaction model and a site response analysis model.

The soil tunnel system model is not convergenced during the static analysis when the soil is transfered into Plasticity. The error is elaborated below:
Thus, I built a simple model, in which the tunnel is replaced with the soil. The site response analysis is implemented in the model. When I tranfered the soil into plasticity. The analysis is convergenced, but another error is raised as below:

Vector::operator/(double fact) - divide-by-zero error coming Why is these two errors raied? Could you give me some advices?

Here is the two numerical models:

The soil-tunnel interaction model: The soil site analysis: Is the material properties is defined right?

Best regards,

Jiaxing